C H A P T E R
26
Aquatic Housing
Jason Cockington
Australia New Zealand Association of Aquarium Professionals inc., Brisbane, QLD, Australia
Introduction to Aquatic Housing
When zebrafish first emerged as a novel genetics
model system, the availability of strategic housing
options was extremely limited. Despite this, the Strei-
singer Laboratory at the University of Oregon was able
to devise a life-support system that enabled the fish to
flourish under their care. The early genetics revolutions
that led to zebrafish becoming today’s second most used
animal model after mice began in common glass
aquaria, purchased from a local pet shop provider and
covered with custom-cut acrylic lids. The fish were
housed in a semiflow-through environment where
water exchange occurred during the day by filtration
of influent mains water through a household carbon
filter and a thermostatic mixing valve; water exchange
was then turned off in the evenings (Lawrence & Mason,
2012).
Today aquatic housing solutions share common
threads with this precursor system, but technology has
made possible many advances, from significant
enhancements in tank design and construction to the
development of smartphone-interfaced, fully automated
electronic water quality monitoring and control instru-
mentation. These advancements allow the luxury of
multiple strategic choices when considering how best
to house this durable model organism.
This chapter will explore available choices in detail,
but before decisions can be made with regards to tank
design, materials selection, system design (single-pass
or recirculating water), or scientific consequences
(such as effects on gut flora), one must first understand
the impact that scale and application have on these
choices, as this is the key driver for determining which
aquatic solution will best enable one’s research goals.
As the zebrafish model continues to break new
ground beyond its traditional biomedical discipline,
many research institutions are making a strategic
commitment to its application by building core aquatic
The Zebrafish in Biomedical Research
https://doi.org/10.1016/B978-0-12-812431-4.00026-9
animal facilities designed to facilitate the complexity of
cross-disciplinary research programs. This commitment
often requires the use of specialized housing equipment,
especially when project designs evaluate the fish over
many weeks or months (Traver et al., 2004). While the
zebrafish has proven itself as an adaptable research
model and a critical core model organism, the institu-
tional commitment worldwide to its use is still very
cautious compared to century-old rodent models. One
would be hard pressed to find an institution with a
biomedical research discipline in its faculties lacking a
rodent vivarium supported by dedicated laboratory
animal technicians, yet the commitment to centrally staf-
fed aquatic facilities is still in an emergent phase. As a
result, it is commonplace for significantly lower scale
zebrafish aquariums to be utilized as the housing solu-
tion for the model. Regardless of the commitment level
toward the zebrafish model system, all facilities strive
to provide the fish with the best care possible within
the restrictions of the aquarium’s design. Thus, housing
systems need to provide the fish with a stable and favor-
able culture condition while also supporting the specific
research goals. In addition, as the application of fish in
biomedical research increases, oversight bodies are
requiring aquatic research facilities to meet and adhere
to more stringent regulatory requirements worldwide
(Lawrence et al., 2009).
When it comes to deciding on the appropriate scale
for a zebrafish commitment, the most influential drivers
are (1) whether multiple research groups are interested
in applying the model, and (2) whether the investigative
questions can be addressed with simple short-term
developmental experiments, or whether longer term,
multigenerational approaches are required. If the desire
for zebrafish application is being driven by only a small
number of investigators who are asking questions that
can be assessed without the need to rear multiple
strains, then the most appropriate housing solution
may be as simple as a glass aquarium tank on a bench.
279 © 2020 Elsevier Inc. All rights reserved.
280 26. Aquatic Housing
If, however, the questions require multiple generations
of controlled pedigree crosses, then the need to provide
husbandry for multiple life stages will require a more
involved commitment to the model, and an expanded
commitment toward the housing technology applied.
This commitment may take the shape of a custom solu-
tion designed and developed specifically for the applica-
tion. However, the most common answer is likely to
come in the shape of a commercially available housing
solution that elegantly combines principles of commer-
cial aquaculture, laboratory animal housing, and
research genetics, and that can be tailored and scaled
to meet the needs of multiple research programs in
one facility.
Beyond defining the application scale, culture condi-
tion is the next most critical element to be addressed in
order to ensure that the fish become a productive
research model. While the major elements of culture
condition will be discussed in other chapters of this
text, there is one element of culture conditions that
requires attention within the discussion of housing solu-
tions. This is stocking density and its impact on stress.
The density at which fish are kept in captivity exerts
profound effects on their health, productivity, and
welfare. Holding densities also ultimately have signifi-
cant implications on the operating costs of an aquarium,
the space required for the housing solution, and the
labor required for maintaining the aquatic facility.
In general, higher stocking densities will ensure that
costs, space, and labor are kept to a minimum, but it
may not be the best approach for achieving a healthy,
productive zebrafish colony.
It is widely accepted that crowding fish at high stock-
ing densities can cause chronic stress, resulting in
changes in growth rate, feeding, behavior, immune func-
tion, and health (Ellis et al., 2002; Procarione, Barry, &
Malison, 1999; Suomalainen, Tiirola, & Valtonen, 2005;
Wedemeyer, 1996). It is also recognized that stocking
density effects on fish welfare are complex and
species-specific; these effects can comprise numerous
interacting factors including water quality, social inter-
action, the carrying capacity of the holding environ-
ment, as well as the special and behavioral needs of
the particular species (Pavlidis et al., 2013).
As a gregarious species, zebrafish benefit from the
ability to form loose shoals with their tank mates, where
they can establish and express social hierarchies. Zebra-
fish are also known to display aggressive behavior asso-
ciated with the establishment of territories and
dominance hierarchies (Larson, O’Malley, & Melloni,
2006); thus, both crowding and low-density housing
are less than ideal for zebrafish. This highlights the
importance of matching holding conditions to their
behavioral tendencies.
III. Husbandry
The currently accepted guideline for establishing a
stocking density policy that is conducive to productive
and healthy zebrafish is based on the approach outlined
by Harper and Lawrence (2011), who suggest that den-
sities of 40e50 fish per liter are appropriate for early
larval culture, but that density should be gradually
reduced to five fish per liter for adults. This commonly
applied strategy was supported by Pavlidis et al.
(2013) who showed that zebrafish held at a density of
10 fish per 2 L of available water volume adequately
expressed normal behaviors while avoiding elevated
cortisol stress reactions; this suggests that even when
water quality is maintained, a minimum water volume
is required for zebrafish welfare.
With this knowledge, it is possible to determine the
necessary scale of the intended application, based on
the target number of animals required to meet the
research program’s needs. With appropriate scale
commitment established, the desired application will
be the key driver for determining the type of enclosures
needed for housing the fish.
Primary Enclosures (Tanks)
In an aquaculture facility, the primary housing enclo-
sures are the tanks. Depending on the scale and applica-
tion of the projects employing the zebrafish model
system, the tanks used will vary in shape, size, and
materials from which they are constructed. They will
also vary in how they handle water exchange and how
they contain the specimens. These factors all affect the
welfare of the fish, the functionality of the life-support
system, and the pace and efficiency of research being
conducted. Thus, aquatic housing choice for any appli-
cation needs to include careful consideration of how a
particular system, and its associated tank types, help
investigators achieve their research and husbandry
objectives.
Materials Selection
Material toxicity is a concern when introducing any
new equipment that will contact the fish. For this reason,
the most common materials that zebrafish holding tanks
are made from are glass, acrylic, fiberglass, polyeth-
ylene, polycarbonate, or polysulfone. These materials
have an established record for use as aquatic housing
solutions, each with its benefits and limitations, and
thus, the appropriate material selection will depend on
the application.
 Primary Enclosures (Tanks)
Thermoplastic Polymers
In modern high-density installations, tanks are most
likely to be made from polycarbonate. This plastic is
durable and relatively inexpensive, and it can be clear
or pigmented. The pigments chosen for zebrafish enclo-
sures are typically shades of blue or green, focused on
filtering out the light spectrum utilized by chloroplasts,
thus reducing the development of algae on tank surfaces
as the aquarium’s water begins to accumulate phos-
phates and nitrates. Polycarbonate tank construction is
made via pressure injection molding, and it is, therefore,
amenable to the production of tanks in a wide range of
shapes and sizes. Most commercial zebrafish aquarium
manufacturers offer a range of polycarbonate tanks,
allowing a significant level of customization to any
aquarium installation. One of the most significant bene-
fits with the use of polycarbonate tanks is that they are
readily movable within the facility. One can thus tempo-
rarily relocate fish to a workbench for investigation,
directly exchange tanks for cage cleaning, or transfer
fish from a fouled tank into a fresh new tank. This
accommodates significant flexibility in cleaning and
disinfection protocols. Although polycarbonate tanks
are lightweight and can withstand mechanical washing
processes, thereby reducing manual labor and zoonosis
risks present with some other tanks, polycarbonate
degrades at high temperatures, and thus, cannot be
repeatedly sterilized. For countering this, several manu-
facturers of polycarbonate tanks also offer a range of
tanks made from polysulfone.
Polysulfone is slightly more costly, with a greyish
hue, and is marketed as having a higher temperature
and chemical tolerance than polycarbonate. Tanks
made from polysulfone can be a suitable substitute for
applications, where regular tank sterilization is
required.
Both polycarbonate and polysulfone tanks are pro-
duced using bisphenol-A (BPA). This is the main draw-
back to the application of these thermoplastic polymers
in aquarium holding tanks, as it has been established
that both types can leach BPA into the water.
Bisphenol-A is an estrogen mimic that can cause serious
reproductive problems in vertebrates, including fish.
Howdeshell et al. (2003) demonstrated that new polycar-
bonate and polysulfone tanks both leach BPA into the
water. However, leaching is significantly higher in
aged polycarbonate tanks, thus elevating this as a poten-
tial concern for established zebrafish facilities. Howde-
shell et al. (2003) also discussed the leaching levels
being similar to those previously reported to skew sex
ratios toward females in the South African clawed
frog, Xenopus laevis (Kloas, Lutz, & Einspanier, 1999),
and that resulted in the presence of ovo-testes and testic-
ular abnormalities, including a decrease in the number
III. Husbandry
281
of spermatozoa, in the medaka Oryzias latipes (Metcalfe
et al., 2001). While definitive effects of BPA leaching in
a zebrafish colony remains to be tested, it is reasonable
to conclude that BPA leaching may have similar effects
on the diminutive minnow. The benefits gained from
the use of thermoplastic polymer tanks are nevertheless
significant, suggesting that regular tank replacement
may be the most reasonable strategy for managing this
side effect until a satisfactory BPA-free substitute mate-
rial can be found.
Glass
If BPA effects are likely to cause issues for the research
program, the inert properties of silica-based glass may
make it the material of choice. Glass tanks are the most
common housing solution implemented in small-scale
operations, as basic rectangular glass aquaria are readily
available at most pet shops. Glass can also be applied in
large-scale operations, and it is the next most common
material used in primary enclosures after thermoplastic
polymer tanks. The principle benefits of glass are that it
is inert, and thus, poses no chemical leaching risk to its
inhabitants; it does not scratch easily, which, in turn,
means it resists the establishment of biofilms and other
tank fouling, and it is relatively inexpensive. Glass,
however, lacks the flexibility necessary for design vari-
ability, making it difficult to apply to as wide a range
of housing solutions as other materials. Glass also has
the significant limitations of being fragile and heavy,
posing major health and safety challenges to personnel
during routine cleaning.
There are health and safety challenges that glass
aquaria present to the personnel who maintain them.
There are manual handling risks associated with mov-
ing, servicing, and replacing the tanks. Large glass
aquaria are heavy, often requiring multiple people to
move and position them. This means they do not readily
lend themselves toward automated washing. As a
result, most glass aquaria will need to be cleaned in
place, often resulting in a laboratory animal technician
arm deep in fish water. While this in itself is not a
dangerous activity, there is a significantly elevated risk
of skin abrasions from this method of tank maintenance,
a risk that is compounded if there are minor chips to the
rims of the tanks. This can prove to be a major obstacle
for facilities that opt to utilize glass for their primary
enclosures, as there is a significant zoonotic disease com-
mon to zebrafish colonies; the atypical tuberculosis,
Mycobacterium marinum (Matthews, 2004). To overcome
this risk, facilities using glass tanks need to be mindful
of their biosecurity efforts, to ensure they do not inad-
vertently contract a mycobacterium infection. Strategies
for maintaining a mycobacterium free facility are
discussed later in this book.
282 26. Aquatic Housing
Acrylic
As the zebrafish model expanded beyond traditional
developmental genetics studies, there grew a need for
specially designed primary enclosures to enable
specific experimental work. As a result, poly(methyl
methacrylate), commonly referred to as acrylic, has
become an important material of choice. Although tech-
nically also a thermoplastic polymer, acrylic has signifi-
cant differences in how it can be applied as a primary
enclosure, and thus, is being discussed separately.
Acrylic does not contain the potentially harmful
bisphenol-A subunits found in polycarbonate and poly-
sulfone plastics, and it is an economical alternative to
polycarbonate when extreme strength is not necessary.
This transparent thermoplastic is often used in sheet
form as a lightweight, shatter-resistant alternative to
glass. It is approximately half the weight of glass, and
while it is significantly stronger and more flexible than
glass, it is also significantly more prone to scratching.
Scratches may be easily removed by polishing the
damaged surface of the material with appropriate sol-
vents available from hobby retailers. Acrylic can be
damaged by organic solvents such as ethanol and by
other chemicals. It also has low heat tolerance, which
can limit sanitation options. Even high-quality acrylic
will begin to disfigure and melt at temperatures above
80
C. Acrylic is very resistant to the effects of strong
alkali solutions, which are extremely effective germi-
cides. Sodium hypochlorite (bleach) can be applied
safely to acrylic tanks at concentrations up to 5%, and
any biofilms or algae can then be subsequently wiped
off with a soft cotton cloth. When using this method of
sanitation, it is necessary to detoxify any residual free
chlorine before repopulating the tank.
Polyethylene and Fiberglass
Although less commonly applied, polyethylene and
fiberglass-reinforced polyester (FRP) also play an impor-
tant role in zebrafish housing. These materials share
similar beneficial traits and are most commonly applied
when the application requires either holding a large
volume of water or controlling illumination to the
culture, thus eliminating the option of using the
optically transparent acrylic.
Emerging Materials
In recent years, 3D printing has significantly
enhanced the ability to construct tools and in this case,
housing solutions. A prime example, the Screen Cube,
is a 3D-printed housing solution that is used for the
transfer of fish between primary holding tanks and
treatment baths, allowing for rapid and cost-effective
chemical screening of adult zebrafish (Monstad-Rios,
Watson, & Kwon, 2017). While 3D printing offers
III. Husbandry
immensely powerful flexibility for creating novel pri-
mary enclosures, it is important to note that not all 3D
printing polymers are considered safe. Macdonald
et al. (2016) assessed the biocompatibility of four
commercially available 3D printing polymers, and via
the observation of key markers in developing zebrafish
embryos, showed all of the photopolymers in an
untreated state to be lethally toxic to zebrafish embryos.
When parts fabricated from one of these photopolymers,
Fototec SLA 7150 Clear, were washed with an organic
solvent (99% ethanol), biocompatibility significantly
improved, making it suitable for use in zebrafish culture.
This is an important message with regards to employing
new materials in a zebrafish aquarium. It is critically
important to run a basic Fish Embryo Toxicity (FET)
test, such as that described by Schulte and Nagel
(1994), to determine the biocompatibility and safety of
employing the new material.
Tank Design
Armed with the knowledge of application and appro-
priate material selection, it is possible to consider tank
design for research program outcomes. Zebrafish tanks
come in a variety of shapes and sizes, defined by their
intended function. Although tank design varies between
vendor and application, all tanks are generally designed
to serve the common purpose of containing the fish
within a specified space and volume of water, while
also maintaining suitable life-supporting culture condi-
tions for its inhabitants.
Box Tanks
Although becoming less common in the modern era,
the traditional box tank still serves an important role for
those who are keen to pilot small experiments but are
unable to access a large-scale production facility. These
tanks are typically 20e50 L in volume, allowing an all
in one approach for the keeping of the fish. The large
water volume allows the fish the space to exhibit their
natural shoaling behavior, while also providing suffi-
cient volume to buffer against shifts in water quality.
The box tank can be used as a freestanding static aquar-
ium, or with some modification, as a flow-through
aquarium, in a semirecirculating life support system,
or in a Recirculating Aquaculture System.
Serial Tanks
Serial tanks are most commonly made from glass and
are applied in large-scale aquariums. Serial tanks are
essentially an advanced modification of a box tank,
resulting in long aquaria that are divided transversely
by partitions to provide multiple compartments for
keeping adult fish (Fig. 26.1). A typical serial tank is
 Primary Enclosures (Tanks)
FIGURE 26.1 Serial Tanks. Serial tanks typically comprise a long
glass box tank, which is divided into housing compartments by par-
titions (black vertical lines). These partitions leave a small gap at the base
of the tank, resulting in a series of interconnected compartments.
Influent water enters the series of tanks in the compartment at one end
and flows through each tank compartment via the gaps at the base of
the partitions (white arrows). Effluent water exits the tank in the last
compartment via a standpipe (passive overflow), setting the water
volume throughout the series. A grill fitted to the head of the standpipe
prevents fish from escaping containment.
120 cm wide, 60 cm deep, and 22 cm high (Brand, Gran-
ato, & Nüsslein-Volhard, 2002). The long box is typically
subdivided by glass plates that leave a small 1.5 mm slit
at the bottom, or via perforated plastic plates, resulting
in a series of 8e10 interconnected compartments of
about 12 L each. Each compartment has a separate lid,
allowing the holding of up to 60 adult fish per compart-
ment at a stocking density of five fish per liter. The
defining feature of serial tanks is that the water in a
row of tanks flows into the compartment at one end,
and then passes through the slits from one compartment
to the next, carrying waste particles with it. The water is
drained from the last compartment at the far end of the
row by overflowing into a standpipe, thus determining
the water level across the row. In order to prevent loss
of fish, the standpipe is typically fitted with a coarse
mesh or grill.
There is potential for capital savings to be made with
the implementation of serial tanks, as there is only one
influent and one effluent point requiring attention for
supporting 10 tanks; however, the fact that serial tanks
are too large to be removed from their support racks
means cleaning must be done in place, making long-
term maintenance of the system inefficient. Thus, the
benefits gained from serial tanks are often at best
neutralized by longer-term inefficiencies.
There is also one significant limitation to consider
before committing to the use of serial tanks. Due to their
interconnectivity, they are not able to control the spread
of horizontally transmitted diseases, as all tanks share a
common water source. Thus, if a population of fish in
chamber one were infected with a disease like micro-
sporidia (Pseudoloma neurophilia), then fish in the other
III. Husbandry
283
nine compartments would be exposed to spores released
into the water as it moves through the serial chambers
toward the outflow pipe. To avoid this problem, the
use of individual, self-contained primary enclosures
may be necessary instead; this is the main reason why
this tank design has lost favor in recent years.
Passive Overflow Tanks
There are many ways a passive overflow tank can be
applied in a large-scale facility, and these tanks offer the
most flexibility in design, shape, and customization. The
simplest implementation of the passive overflow tank is
described by Brand et al. (2002), in which individual
containers (plastic or glass) are placed in a row on to a
bunded shelf. Water enters into each tank through an
influent hose or pipe and exits through an overflow
point on to the shelf, from which it is drained into a com-
mon water treatment unit. However, this implementa-
tion of the passive overflow tank design often results
in fish waste accumulating around the base of the tanks,
and thus, more frequent cleaning is needed. An alterna-
tive implementation of the passive overflow tank
utilizes a standpipe or tank flange, which allows the
overflow point of the tank to couple to a plumbed waste
line, keeping all effluent water contained within the
wastewater drainage pipes. Both strategies typically
rely on fine mesh grills, or filter media, at the overflow
point to avoid fish escaping containment, and this can
be tailored to suit adult or juvenile fish. Examples of
both strategies are illustrated in Fig. 26.2.
Passive overflow is the most common strategy
applied to specialty application tanks when they are
incorporated into system solutions rather than existing
as stand-alone housing chambers. Some of these solu-
tions to unique applications will be discussed
subsequently.
Active Overflow Tanks
Despite the significant flexibility advantage passive
overflow tanks offer, the nature of the hydrodynamics
within the tanks means these solutions offer limited abil-
ity to maintain tank hygiene, and frequent cleaning
intervention is needed to keep these tanks functioning
as intended. To overcome this challenge, it is common-
place for large-scale aquariums to implement the use
of solids-lift (active) overflow tanks as their principle
housing device. In these solutions, the tanks employ
the use of either siphons and/or specially designed par-
titioning baffles to alter the hydrodynamics within the
tank, such that an undertow current is established,
creating an active draw across the base of the tank
(Fig. 26.3). Providing there is an adequate velocity of wa-
ter entering the tank via the influent pipe, the undertow
284 26. Aquatic Housing

FIGURE 26.2 Passive Overflow Tank Application. (A) In this schema, individual holding tanks are placed onto a bunded shelf. Water enters
into each tank via a small outlet from a header pipe manifold that runs along the length of the shelf. Water exits the tank through a grill, or series of
small holes, at the top of the tank. Suspended particulate waste (fines) trickles down the tank onto the shelf (silt stream). Wastewater pools on the
bunded shelf, and is collected into a common drainpipe that sequesters the wastewater to the filtration assembly. (B) Passive overflow, glass box
tanks coupled to a wastewater drainpipe via tank flanges (C) Passive overflow, polycarbonate tanks on a bunded shelf.

current will have sufficient strength to carry waste
particles out of the tank.
Specialty Application Tanks
Beyond the range of designs that make up the hous-
ing tanks, there are a number of unique primary enclo-
sures designed to facilitate specific application needs,
the most common enabling controlled breeding. Mating
tanks are a critical element of an aquarium supporting
research projects that focus on developing zebrafish
embryos.
Mating Tanks
Historically, adult zebrafish were encouraged to
spawn their eggs via a process known as “marbling,”
where an investigator would cover the entire bottom
of the housing tank with a layer of marbles, effectively
creating a series of nooks and crannies that allow fertil-
ized eggs to drop beyond the reach of hungry tank mates
(Westerfield, 2000). The fertilized embryos would then
be siphoned from the tank every morning. While effec-
tive, marbling lacks the finesse that is required to meet
the productivity demands of many research aquariums
today; hence the development of purpose-designed
breeding tanks.
Mating tanks come in various shapes and sizes,
ranging from just over half a liter up to 2L volumes
(Fig. 26.4). These devices can be applied directly in the
main housing tanks; however, are most commonly
applied as short-term, static enclosures, housing a pair
of fish, or in the larger cases, small groups of fish, based
on a technique that was first described by Mullins,
Hammerschmidt, Haffter, and Nüsslein-Volhard
(1994). They all share the same fundamental design
characteristics: a tank to hold the water, a trap to sepa-
rate the adult fish from their spawn, a divider to segre-
gate the adults by sex and so enable timed mating,
and a lid to prevent the fish from jumping out of the
tank. While there is some handling stress associated
with the use of these tanks, their independence from
the main housing enclosures gives them significant
advantages for both the harvesting of the embryos and
for sanitation after use.

III. Husbandry
 Primary Enclosures (Tanks)
FIGURE 26.3 Active Overflow Tanks. (A) This shows the basic
hydrodynamic design of a solids-lift tank. Water is delivered (blue line)
into the tank toward the front of the tank. The unique design of the
partitioning baffle (or siphon) at the back of the tank results in a
pressure differential between the holding space and the overflow.
A strong undertow current (red arrows) draws particulate waste toward
the back of the tank and lifts it up to the tank overflow. (B) Examples of
typical active overflow tanks.
FIGURE 26.4 Mating Tanks. Examples of various polycarbonate
mating tank designs.
For many years, the vast majority of mating tank
designs followed a simple, flat-bottom approach to the
trap, which although satisfactory to yield sufficient
quantities of embryos from the fish, did not play to the
fish’s natural spawning site preferences. Following
research conducted by Sessa et al. (2008) that illustrated
a preference to spawn in the shallows, as evidenced by
the observation that the zebrafish actively chose to
spawn at the shallow end of a sloped trap, many
III. Husbandry
285
aquariums began to establish gradients in their mating
tanks by placing them on a sloped platform or by
wedging the trap with the divider, such that one end
sat higher in the tank. The widespread adoption of this
behavior led to one commercial manufacturer refining
its mating tank design, resulting in a beach style trap,
with a sloping layered surface, that allows the fish to
select their preferred depth upon it for spawning.
Mass Embryo Production Systems
While the above mating tank designs are effective,
there are scenarios where their application cannot easily
meet the production demand for fertilized embryos.
When obtaining large quantities of developmentally
synchronized embryos is more important than tracking
pedigree of individual contributors, Mass Embryo
Production Systems (MEPS) offer a unique solution.
The main impetus for shifting breeding tank design to
capitalize upon the group spawning dynamics of the
zebrafish came as a result of the model expanding into
the realms of toxicological studies, chemical screens,
and other applications that demand increasingly large
embryo numbers. Adatto, Lawrence, Thompson, and
Zon (2011) describe placing up to 300 fish inside a
custom-designed, 100 L static group spawning chamber
capable of yielding 10,000 fertilized embryos within
10-minute harvesting windows.
There are several different models of MEPS currently
available commercially (Fig. 26.5), each capable of use as
a static breeding tank, or of being coupled to an aquar-
ium system as a passive overflow device. One of the
most significant differences between the MEPS solutions
from various commercial suppliers is the base material
with which these tanks are constructed. One design is
completely transparent, constructed from acrylic, allow-
ing the tank to leverage the existing facility
FIGURE 26.5 Mass Embryo Production Systems (MEPS). Large
numbers of broodstock fish can be housed and spawned directly inside
these devices, facilitating on-demand collection of synchronized
embryos. From left to right, an example of a polyethylene MEPS, large
(60 L, 200 fish) and small (13 L, 40 fish) acrylic MEPS, and a fiberglass-
reinforced polyester MEPS. Courtesy of Aquaneering Inc., Courtesy of
Tecniplast Aquatic Solutions and Courtesy of Pentair Aquatic Eco-systems.
286 26. Aquatic Housing
photoperiods; another has opted to construct their
MEPS solution from polyethylene; while a third supplier
chose to fabricate their tank using FRP. The latter de-
signs both allow photoperiods to be established inde-
pendent of the main facility light cycle. Other
differences between the various MEPS include size,
holding capacity, spawning platforms, mode of egg
collection, and patterns of water circulation; but regard-
less of design, the scale and reliability of function of
these tools are tightly coupled to scientifically grounded
management of the broodstock fish placed inside (Law-
rence, 2011).
Respirometers and Mazes
While traditionally the model’s strength lay with the
developing embryo in biomedical discoveries, and so
too with the majority of specialty application tank solu-
tions, the adult zebrafish has started to become a more
significant model to other disciplines; and with this,
new applications for the model have emerged. Swim-
tunnel respirometers, puzzle mazes, and operant condi-
tioning shuttle boxes have all emerged as significant
tools in the zebrafish toolkit in recent years.
One such tool, the respirometer, has gained particular
focus for investigators interested in the physiology of
FIGURE 26.6 A Brett-style Swim-tunnel Respirometer can be used
for measuring the physiology, energetics, behavior, biomechanics, and
kinematics of swimming fish. (A) Overhead schema: (1) variable speed
motor, (2) propeller shaft, (3) propeller, (4) baffles to assist with
achieving laminar flow, (5) overflow pipe, which extends above water
surface, (6) sealable port for oxygen sensor, (7) honeycomb grid to
assist with establishing laminar flow, (8) working section where fish are
housed, (9) water bath, (10) water inlet, one-way flow (11) water bath
passive overflow (B) Swim-tunnel respirometer. Courtesy of Loligo
Systems.
III. Husbandry
the zebrafish, examining traits such as swimming speed,
stamina, oxygen demand, and body shape drag. The
most commonly applied, Brett-style Swim-tunnel Respi-
rometers (Fig. 26.6) have a rounded-rectangular shape
and contain a propeller that drives water circulation in
one direction, while the fish are restricted to a long
compartment known as the “working section,” where
the water passes in a laminar flow profile (Brett, 1964).
A variable-voltage motor typically drives the propeller
in the swim-tunnels, such that a range of water speeds
can be obtained. These devices are typically constructed
from acrylic, allowing optical transparency for simpli-
fying data collection, and can be applied statically, fitted
with a hang-on filter or connected to an aquarium
system as a passive overflow tank. Respirometers can
be applied as open swim-tunnels for projects such as
examining gross morphology impact on stamina
(Conradsen, Walker, Perna, & McGuigan, 2016); but
when they are used for determining physiological
processes, such as oxygen demand in exercising fish, it
is necessary to seal the experimental chamber from the
surrounding water bath to prevent water mixing, so
that precise measurements can be achieved. In this situ-
ation, a flush pump should be used to refresh the water
within the treatment chamber at regular intervals, to
ensure the dissolved oxygen levels within the respirom-
eter remain above 80% saturation at all times to mini-
mize the chance of measurements being influenced by
hypoxia-induced metabolic adjustments (Clark, Sand-
blom, & Jutfelt, 2013).
Mazes and shuttle boxes have also made inroads as
specialty application tanks for zebrafish in recent years.
Employing the use of a shuttle box, Sneddon (2011)
demonstrated that when housed in an environment
with the choice of a barren or structurally enriched con-
dition, zebrafish exposed to a painful acid stimulus will
override their preference for the enrichment and self-
administer analgesics offered only on the barren side.
Environmental Enrichment
One common trait across most of the tank designs
discussed thus far is that they are designed to operate
as a barren enclosure. This has been a strong barrier
against the use of structural enrichment in zebrafish pro-
duction facilities. In addition, it remains unclear
whether structural enrichment, designed to mimic the
habitat complexity that wild zebrafish can experience,
also benefits laboratory strains, particularly when such
interventions are balanced against costs, such as the
increased difficulty of observing fish and the increased
accumulation of mulm (detritus), which in itself can
lead to morbidity and mortality (Parichy, 2015).
 Secondary Enclosure Systems
While there is mounting evidence suggesting that
zebrafish may indeed benefit psychologically from
structurally enriched environments (Collymore,
Tolwani, & Rasmussen, 2015; Manuel et al., 2015;
Maximino, De Brito, de Mattos Dias, Gouveia, & Morato,
2010; Spence, Magurran, & Smith, 2011; Weber & Ghorai,
2013), little work has been done to address the concerns
relating to colony-wide health impacts that may emerge
as a result of changes in tank hydrodynamics, and the
consequential diminished tank hygiene due to
the presence of alien structures within the water body.
Also, the material selection needs to be carefully consid-
ered when opting to implement the use of a structural
enrichment device to ensure there is no leaching of
unwanted chemicals into the water column. Fig. 26.7
illustrates how some facilities are implementing the
use of structural enrichment.
While it is still relatively uncommon to utilize struc-
tural enrichment in large-scale production facilities,
FIGURE 26.7 Structural and Social Enrichment. (A) A small PVC
T-joint provides a tunnel refuge for subordinate fish. (B) The Zeb de-
vice, a structural enrichment solution offered by Otto Environmental,
provides a floating canopy designed to mimic the dense vegetation of
the native habitat of the zebrafish. (C) For avoiding the detrimental
effects of housing zebrafish in low densities, important carrier pairs
can be housed with phenotypically distinct zebrafish, allowing them to
form social shoals and mediate their aggression by establishing
dominance hierarchies with the other fish. (B) Photo courtesy of Lucie
Nedved.
III. Husbandry
287
it is commonly accepted that it can be significantly detri-
mental to house individuals in isolation (Fox, White,
Kao, & Fernald, 1997; Harris & Bird, 2000; Larson
et al., 2006; McCarthy, Carter, & Houlihan, 1992).
In low-density situations, such as keeping a carrier
pair segregated, fish can exhibit enhanced aggression
and will often fight until one fish either escapes the
conflict by jumping out of the tank or is harassed to
the point of death. This is a major concern for ensuring
the longevity of critically important animals.
In a situation where structural enrichment may not be
readily available for providing refuge for the subordi-
nate, or a territory for the dominant to defend, the
simplest way to manage the aggressive behavior of the
fish is through social enrichment. As discussed previ-
ously, zebrafish have a preference to form small shoals.
This behavior allows a more balanced social hierarchy
to form, and this can be easily established by combining
phenotypically different strains of fish. For example, if a
carrier pair has been identified on an AB background,
adding a small number of TL or TLN strain fish can
preserve the social welfare of all fish in the tank. This al-
lows the fish to maintain a natural social hierarchy,
while also being easily identified as the specimens of
interest to the investigator.
Secondary Enclosure Systems
While in a small-scale operation, the tank may serve
as both the primary and secondary enclosure, but in
large-scale production facilities, the zebrafish tend to
be housed in relatively small primary enclosures, to
maximize the number of specimens contained within a
workspace. If the water within these tanks is not
constantly cycled and cleaned, the fish will quickly suf-
fer from the toxicity of their metabolic wastes accumu-
lating in the water. For overcoming this, the tanks are
integrated into much larger secondary enclosures,
making up dedicated life-support systems designed to
allow influent fresh water to replace the waste accumu-
lating effluent water.
Static Enclosures
The challenge with static enclosures is that from the
moment fish are added into the tank, the life-support
condition, specifically the water quality, begins to deteri-
orate. As the fish respire, they consume dissolved
oxygen and release carbon dioxide and ammonia into
the water. As these metabolic wastes accumulate in the
water, the environment becomes less favorable for its
inhabitants. For this reason, static tanks need to be
kept at significantly lower densities than enclosures in
288
which constant water exchange is possible, and thus, the
situations where static enclosures are appropriate
should be carefully considered.
A significant advantage of static enclosures is that
they avoid all costs associated with establishing and
servicing filtration assemblies, and are, therefore, very
cheap to implement. However, because these devices
lack the ability to clean the water, in comparison to other
life-support solutions, they require significantly more
space per capita. An additional consequence of avoiding
the complexity of filtration assemblies is that static
enclosures require high labor commitment in order to
keep the life-support condition within suitable parame-
ters for the fish.
As a general guide, if the fish are to be maintained
long term in a static enclosure, there are a few basic
tips that will help improve the success of the colony.
To keep water degradation within manageable toler-
ances, Brand et al. (2002) advise densities in static tanks
should not exceed two fish per liter. Only feed the fish to
satiation, so that uneaten food does not amass in the
tank, releasing ammonia and rapidly degrading the
water quality. Allow the fish opportunity to digest and
defecate before performing a tank clean and water
change. Keep a small sponge, or other porous material,
in the tank to act as a basic biofilter. Even without aera-
tion, these measures will assist with the natural biolog-
ical process associated with converting toxic ammonia
into the significantly less toxic, oxidized state (nitrate).
Also, where possible, aerate the water to increase gas
exchange, removing dissolved carbon dioxide and
regenerating dissolved oxygen.
On a more temporary timeframe, adult fish can be
comfortably maintained in static tanks for up to 2 weeks
at low densities without water exchange if feeding is
withheld and the fish are prevented from stressful inter-
actions, such as breeding or fighting. This is an ideal
strategy to exploit when genotyping individuals and
FIGURE 26.8 Hang-on Filter. With the application of a hang-on filter to an otherwise static holding tank, the useable lifespan of the culture
water can be significantly extended. A small powerhead pump draws water up and out of the tank. The water then passes through a basic
mechanical filter pad, activated carbon filter, and biological filter sponge before freefalling back into the holding tank, offering an opportunity for
gas exchange.
26. Aquatic Housing
can be easily achieved by placing two fish in a 1-L
mating tank with a divider separating them.
As zebrafish alevin (yolk sac larvae) transition into fry
(exogenous feeders), they are also often kept in static
enclosures for the first few days of their foraging exis-
tence. As described by Best, Adatto, Cockington, James,
and Lawrence (2010), first-feeding zebrafish fry can
thrive at a density of 50 fish in 200 mL (250 fish per liter),
highlighting that the challenges associated with static
enclosures are also life stage-dependent. By far the
greatest application for static enclosures, however, is
their use in spawning, as was discussed earlier in this
chapter.
Before moving onto the more sophisticated secondary
enclosure solutions, it is worth noting that the addition
of a basic hang-on filter can significantly enhance the
life-support condition of a static tank. By adding one
of these devices to the side of a tank, a semirecirculating
system is achieved, in which the main body of water
within the aquarium is pulled out of the tank by a small
powerhead pump, and then passes through a basic me-
chanical filter (usually a pad akin to the scotch scourer)
and a small chamber containing extruded activated car-
bon. The water may also pass through a more porous
sponge-like material that acts as a biological filter before
the water flows back into the tank, encouraging agitation
and enhancing gaseous exchange at the water surface
(Fig. 26.8). The principal benefit achieved by the use of
a hang-on filter, is the minimization of mulm, the sludge
that collects at the bottom of an aquarium, consisting of
fish fecal pellets, decaying food matter, and other
assorted detritus. By keeping a steady water turnover,
the hang-on filter is able to second waste particles into
the filter before they have a chance to settle in the base
of the tank, significantly extending the useful life of
the water within the tank.
III. Husbandry
 Secondary Enclosure Systems
Recirculating Aquaculture Systems
Recirculating Aquaculture Systems (RAS) are the pre-
dominant secondary enclosure solution applied in
zebrafish facilities of all sizes. The distinguishing feature
defining RAS aquaria is that the water used to house the
fish is cleaned and reused, minimizing water consump-
tion and waste. This allows the total water volume of a
RAS aquarium to be shared by a number of primary
enclosures, thereby, greatly reducing the space required
for housing the research colony. Mastering the process of
cleaning the recirculating water is the greatest challenge
for applying a RAS solution, and its implementation can
vary significantly from one aquarium design to another,
but at its core are five key elements: mechanical filtra-
tion, biological filtration, gas exchange, chemical filtra-
tion, and disinfection.
Mechanical filtration is typically the first step of a
RAS filtration assembly and targets the isolation and
removal of suspended solids from the effluent water.
Solid waste enters the system as a result of uneaten
food or fecal pellets and must be quickly removed
from the circulating water because as the waste decom-
poses, becoming a food source for heterotrophic bacte-
ria, dissolved oxygen is significantly reduced and
ammonia is released into the water column. Particle fil-
ters used for mechanical filtration may be either depth or
surface filters that trap solids as the water flows through
it.
With the solid carbonaceous waste eliminated, the
next phase of water treatment typically focuses on bio-
filtration. Biological filtration describes a process har-
nessing the power of chemolithotrophic bacteria that
oxidize ammonia into nitrite, and then nitrate, to
detoxify the nitrogenous waste. Biofiltration is a highly
aerobic process, requiring double the ammonia concen-
tration in dissolved oxygen within the reaction chamber
(Chen, Ling, & Blancheton, 2006). For this reason, the
biofilter is often coupled with gas exchange (oxygen
in, carbon dioxide out).
Oxygen is a staple for aquatic life, just as it is for
terrestrial life; therefore, as water circulates through
the holding tanks, oxygen will be consumed, and carbon
dioxide added into the water column in its place. If the
active replacement of the water in the holding tanks is
insufficient, dissolved oxygen will be depleted, and car-
bon dioxide concentrations will rise to stressful levels,
diminishing the life-support characteristics of the hold-
ing tank. A minimum of three to four tank changes per
hour is generally recommended (Baensch & Riehl,
1997), with most aquarium operators aiming for at least
five changes per hour through each tank.
Some aquariums employ the use of macroalgae or
additional filtration elements to further digest the
III. Husbandry
289
nitrogenous waste, filtering nitrate out of the water
also. However, this is not common practice within zebra-
fish facilities at this time, with most opting to perform
regular water exchange to keep the nitrates from rising
to dangerous levels. While the precise measure of water
exchange will depend on the rate nitrates accumulate in
the water column, most facilities apply an exchange rate
of 10% over 24 h.
Following the biofilter, the water passes through
chemical filtration, focused on removing colloidal solids
and dissolved organic compounds (DOCs). While not
essential for life support, this additional filtration step
significantly enhances the clarity of the water, which is
essential for achieving ultimate efficiency in water disin-
fection if an ultraviolet (UV) filter is employed.
There are many different technologies available to
achieve these five key elements; thus, a significant factor
driving the underlying decisions that shape the filtration
assembly of a RAS aquarium is the application scale and
system volume.
Small to Medium Scale RAS Aquaria
When a research program expands beyond one or
two production tanks, independently maintaining the
individual tanks becomes very labor intensive and
impractical as a viable option for maintaining the colony.
The hang-on semicirculating filtration solution is inade-
quate as this can only service the tank it is directly
coupled to, and thus, a true RAS solution is necessary
to provide optimal life support for the fish in a cost-
effective manner. Most manufacturers of research
aquaria enter into the market at this level, designing
small RAS solutions affectionately referred to as
“Bench-Top” or “Stand-Alone” systems depending on
size.
Stand-alone systems typically take the form of a rack,
resembling a library bookcase, filled with tanks on the
top section, and with a compact filtration assembly
condensed into the space immediately below the tank
holding space (Fig. 26.9). The rack provides the struc-
tural support for the tanks and the necessary plumbing
to deliver and remove water. The rack is typically con-
structed from marine grade stainless steel, and it is
divided into several presentation rungs. The exact size
and shape of the rack will vary between manufacturers,
with the smallest “bench-top” solution comprising two
rungs of 10 tanks, and the larger stand-alone racks
comprising up to six rungs of 10e20 tanks. A typical
space footprint for a stand-alone system would be
1620 mm  565 mm  2360 mm but will vary with
manufacturer.
The dominant feature of a stand-alone aquarium is
the water treatment unit (WTU) residing directly
290 26. Aquatic Housing
FIGURE 26.9 Stand-Alone RAS Aquarium. A single rack of tanks,
equipped with an integrated water treatment unit, enables a space-
efficient housing solution for a small colony of zebrafish. Effluent
wastewater flows through a prefiltration device as it enters the tech-
nical sump, where the buffering water volume resides, and the culture
condition is monitored and manipulated. The pump draws the water
from the technical sump, driving it through the mechanical and
chemical (carbon) filters, and through the disinfection chamber before
cycling the water up the vertical riser into the header pipes and then to
the fish tanks.
under the bottommost rung of the system. Wastewater
enters the WTU by gravity via a downpipe, typically
passing through a coarse mechanical filter, often referred
to as the prefilter. The water then accumulates in a tech-
nical sump, where essential culture conditions such as
temperature, pH, and conductivity are manipulated. A
pump then drives the water through the filtration as-
sembly, up a vertical riser, and to the tanks via header
pipes and taps. The water then passes through the tanks,
exiting via passive or active overflow into gutters that
connect into the downpipe, thus closing the circle.
Large Scale RAS Aquaria
When the aquarium application demands a larger
scale operation than can be achieved with a stand-
alone rack or two, the most common solution is to
combine several racks together with a central filtration
assembly. These multilinked systems offer the advan-
tage of high holding capacity with enhanced space
efficiency at a relatively low operating cost. While the
basic elements defining RAS aquaria remain in large-
scale production facilities, there are additional elements
III. Husbandry
to consider that impact the function of large multilink,
central life-support (CLS) aquariums.
The greatest benefit that comes with a multilinked
CLS aquarium is that hundreds to thousands of
relatively small tanks can reside on a single secondary
enclosure; thus, maintenance only needs to be
performed on one life-support system making the labor
contribution per tank significantly lower than other
aquarium designs. There is, however, a significant trade-
off that comes with this gain in efficiency, and it must be
weighed against the risks to the research application
before its suitability can be accepted. With one
life-support system supporting so many tanks,
biosecurity in the aquarium is a significantly enhanced
challenge, as a disease outbreak can rapidly circulate
to the other tanks and animals residing in the common
recirculating water. Therefore, as more racks are
committed to a multilinked CLS aquarium during the
design phase of the facility, it is essential to ensure that
adequate capital is allocated to the filtration assembly
in order to ensure maximum biosecurity for the system.
As a guide, the “Plus 10” approach is an excellent means
of ensuring sufficient CLS capability. This approach
suggests that if the final RAS aquarium will consist of
50 racks, and 18,000 L of water, then the CLS should
be designed to support the biological load of a 55 rack,
20,000 L system, or 10% more than the maximum
carrying capacity of the aquarium. This will ensure
that even if the aquarium is pushed to capacity for an
extended period of time, water quality and
life-support condition will not be compromised.
As the scale of the aquarium is considered, it is critical
to ensure sufficient water volume is available in the CLS
to buffer against water chemistry changes in the system.
For example, as the fish respire, and as the biofilter
converts ammonia waste, the water in the aquarium is
acidified as a result of an introduction of carbon dioxide
and hydronium (Hþ) into the water column respectively.
If there is not sufficient hydro volume in the aquarium,
this change will progress quickly, making it difficult to
manage appropriate chemical additions into the water
that would be necessary to compensate the culture
condition change.
Currently, there are two common strategies for
ensuring a sufficient volume of buffering water in a
zebrafish RAS aquarium: a single large technical sump
or a series of multilinked sumps that reside under the
main holding racks. While both methods are effective,
each has its limitations.
The under-rack sump option is an ideal solution
when there is little flexibility with regards to useable
space for the aquarium (Fig. 26.10). By storing the buff-
ering water in the space below the tank holding space of
a rack, it is possible to store the majority of the aquar-
ium’s volume within the tank room, requiring only a
 Secondary Enclosure Systems
FIGURE 26.10 Under-rack Sumps. A bank of four holding racks,
fitted with under-rack sumps.
small additional space for the filtration assembly and
technical sump. As a result, it is possible to house signif-
icantly larger systems in relatively small facilities.
The most significant drawback of the under-rack
sump solution is the fact that the main buffering volume
of the aquarium is stored as “dirty” water. As waste-
water leaves the tanks and enters the rack’s drainage
system, the water flows into the downpipe and typically
passes over a coarse prefilter pad before it accumulates
in the under-rack sump. The water does eventually
pass into the filtration assembly and is cleaned before
being reused in the housing tanks, but the wastewater
that accumulates in the under-rack sumps moves signif-
icantly slower than the rest of the aquarium’s water flow,
and as a result suspended solids (fines) settle out and
begin to accumulate in the base of the sumps. While
291
this accumulated mulm itself is not necessarily harmful
to the aquarium system, it can become a food source for
a plethora of micro-fauna, including potentially patho-
genic bacteria. The accumulation and digestion of
mulm also result in significant inefficiency within the
biological system of the aquarium, as heterotrophic
bacteria compete with the nitrifying bacteria for
resources, such as alkalinity and dissolved oxygen
within the water as they feed on the mulm. This side
effect of leveraging an under-rack sump solution is
readily managed by regularly servicing the sumps.
If the aquarium is fitted with a disposable depth filter
(sock), this service may be as simple as agitating the
sumps to re-suspend the accumulated mulm so that
the mechanical filter can remove the fines. This should
obviously be performed prior to a scheduled filter ser-
vice. If the aquarium is instead fitted with a surface filter,
such as a drum filter, then some of the solid waste parti-
cles will be finer than the filter screen and will simply
pass through the filter. In this instance, siphoning the
mulm from the bottom of the sumps would be an appro-
priate means of managing the situation.
If there is more space available for the CLS when the
aquarium is being designed, it is possible to avoid the
complications of under-rack sumps by increasing the
size of the technical sump to accommodate the buffering
water volume. Allowing opportunity for increasing the
depth of the sump this large central sump can either
be sunk into the floor of the facility, or it can simply
reside as a large above-ground holding tank (Fig. 26.11).
The main benefit that comes from a central sump
design is that it allows the buffer water to be stored in
a somewhat cleaner state. As wastewater leaves the
rack, the downpipe directs it into another wastewater
pipe that channels the water to the filtration assembly,
then typically, directly into a first stage mechanical filter

FIGURE 26.11 CLS Sumps. Large central life-support sumps can be sunk into the floor of the facility (A), increasing the maximum depth of the
sump, effectively holding the system’s buffering water volume vertically. When this option is not available, the CLS sump can reside as a large
above-ground holding tank but will take up significantly more floor space as the water volume is held horizontally (B), unless the system racks are
at a significantly higher elevation. Photo B courtesy of Glen Turner.

III. Husbandry
292 26. Aquatic Housing

or dirty sump. Thus, the water can be processed by the
filtration assembly prior to being “stored,” limiting the
volume of water capable of supporting unwanted micro-
biota. While this solution allows the first stages of the
RAS process to complete (removing solid wastes via
mechanical filtration, converting nitrogen wastes via
biofiltration, and possibly even stripping the water of
DOCs and other colloidal solids in the chemical filter),
it is prudent to ensure that the disinfection step always
occurs immediately prior to the water being returned
to the holding tanks, as effective disinfection is transi-
tory in a recirculating system.
Water Loop
With buffer volume addressed, the next major
decisions will center on the recirculation of the water.
A major challenge in designing a large-scale zebrafish
aquarium is being able to manage the dynamic pressure
changes that occur within the aquarium as depth filters
foul and as more weep points emerge in response to the
opening of taps as tanks are added to the system. These
challenges are addressed in the water loop strategy of
the aquarium.
A dual-loop system, similar to that operating at ZIRC,
the Zebrafish International Resource Center (Varga,
2011), requires the use of multiple pumps to circulate
the water through the aquarium and is typically
employed when the filtration assembly is not able to
rely on gravity pressure to drive the mechanical filtra-
tion step. In this scenario, the aquarium will typically
drain into a wastewater collection sump from which
the filtration pump will draw water. The dirty water is
driven through the filtration assembly and returns into
a clean sump. The clean sump is typically fitted with
an overflow that returns water to the dirty sump if
unused. This is the first loop. The second loop is driven
by the circulation pump, which drives water from the
clean sump via a disinfection chamber, through the
header pipes to the tanks. The water then drains from
the holding tanks into the dirty sump. The circulation
loop is also typically fitted with a rack bypass, allowing
water to be diverted directly into the dirty sump when
the holding racks are sparsely populated. An example
of a dual-loop system is illustrated in Fig. 26.12.
The benefit of a dual-loop system is that it allows
greater flexibility in filtration assembly design, while
also ensuring that tank circulation is not impacted by
changes in the hydrodynamics of a fouling filter. Dual-
loop systems also have a significant disadvantage,
however, which is the need to balance the pumps.
If the circulation pump has a greater draw from the clean
sump than the filtration pump is delivering, water level

FIGURE 26.12 Dual-loop RAS aquarium. The Filtration Loop (1) processes the effluent (used) water, making it suitable for reuse. Used water
drains from the holding tanks and is collected in the Dirty Sump. The used water is pumped through the mechanical filtration assembly, elim-
inating the large particulate wastes. The water is then cycled through an activated carbon chemical filter, removing the fines and colloidal solids,
including dissolved organic compounds, the carbonaceous wastes that yellow the water and act as a food source for heterotrophic bacteria. Finally,
the water passes over a trickle filter before entering the Clean Sump. The trickle filter serves as both a biofilter and gas exchange column.
Denitrifying bacteria populate the oxygen-rich filter, converting ammonia waste into less toxic nitrite and nitrate. Water from the Clean Sump can
reenter loop one by overflowing into the Dirty Sump, or it is pumped into loop 2 as required. The Circulation Loop (2) pumps water from the Clean
Sump through the UV disinfection chamber to the header pipes servicing the tanks in the fish room. In the fish room, water passes through the
holding tanks, emerging as used water entering a central wastewater drain and returns to the Dirty Sump (Loop 1). Loop two also contains a rack
bypass, allowing water to be cycled directly from the Clean Sump to the Dirty Sump, reducing rack pressure of a lightly populated fish room, and
simplifying the pump balancing process.

III. Husbandry
 Secondary Enclosure Systems
will begin to drop, and could lead to the circulation
pump pulling air into the water as a result of a vortex
forming near the pump pickup, which in turn, could
result in a Gas Bubble Disease (GBD) event.
For any given temperature, salinity, and barometric
pressure, there is a maximum stable level of dissolved
gases within the water. When this is exceeded, the total
gas pressure (TGP) is in an unstable, super-saturated
state. In a recirculating aquarium system, the TGP will
remain supersaturated while in the header pipes,
because the water is under pressure; however, the
dissolved gases begin to release from the water as
soon as it is freed from the pressurized condition. In a
RAS aquarium, this occurs in the holding tanks,
exposing the fish to a potentially lethal situation.
In extreme cases, it is possible to observe tiny bubbles
effervescing out of solution; however, GBD is usually
far subtler, manifesting as microscopic bubbles in the
soft tissues of the fish, leading to death from asphyxia-
tion following ischemic necrosis of gill lamellae, as the
air bubbles disrupt the blood flow (Pauley & Nakatani,
1967; Weitkamp & Katz, 1980).
There are several options available to control the risks
associated with GBD and pump imbalance. A TGP
sentinel probe offers a safeguard against GBD by allow-
ing the aquarium system to have an automatic cutoff
trigger. In a dual-loop aquarium, the TGP sentinel probe
is coupled to the circulation pump and monitors the
TGP in the header pipes. If the TGP rises above 100%,
a relay is triggered that cuts power to the circulation
pump. The filtration pump can continue to circulate
without risk to the fish, depending on system design.
Although the fish are now at risk from complications
due to the lack of circulation, the lethal effects are signif-
icantly slower than GBD, giving the operator an oppor-
tunity to address the fault. While effective, TGP sentinel
probes require regular maintenance and calibration to
ensure they are functioning correctly.
To effectively protect the aquarium from GBD, it is
important to measure TGP and not rely solely on shifts
in dissolved oxygen as a guide for GBD risk, as nitrogen
is a significantly greater percentage of the atmosphere.
It is thus, possible to experience a GBD event when
dissolved oxygen is below saturation.
While there is no better substitute available for safe-
guarding against GBD, in a dual-loop aquarium it is
prudent to also install level sensors in the sumps that
allow an operator to shut down pumps to ensure there
is no risk of drawing air into the system if the pumps
get out of balance. In some cases, an aquarium
life-support computer can also manage balancing, if
the system is fitted with Variable Frequency Drive
(VFD) pumps.
A VFD is used for adjusting the flow of a pump to the
system’s actual demand by controlling the frequency of
III. Husbandry
293
electrical energy running the pump. It is most typically
applied on aquariums that employ larger three-phase
pumps. However, recent advancements in this technol-
ogy have allowed the development of VFDs also to
control smaller single-phase pumps. Aside from allow-
ing enhanced control over a pump’s performance,
VFDs also allow the pump to constantly run at peak
efficiency, by matching the pump curve to the system
curve.
In a single loop, or dead-leg system, the challenge of
pump balancing does not exist, but there are other chal-
lenges coupled with relying on a single drive point to
maintain both filtration and circulation. As with the
dual-loop aquarium, a dead-leg system will experience
significant shifts in the hydrodynamic resistance as
filters foul, and tanks require more available water
flow. As a result, VFD pumps are a staple element of a
single loop aquarium.
In order to achieve sufficient flow through every tank
at maximum holding capacity, the pump needs to be
capable of delivering sufficient flow velocity to over-
come the resistance that mounts as more taps are opened
and as filtration media fouls. In addition to these
obvious sources of resistance, it is important to
remember that zebrafish aquarium system curves have
a significantly greater resistance than other aquaculture
facilities of comparable size, because of the number of
header pipes carrying water to the tanks, and the narrow
diameter of the final outlet from which the water must
pass into the tank. Without the sophistication of VFD
pumps, a large dead-leg zebrafish aquarium would
need to either run the system at capacity all the time,
creating a significant burden from the need to clean
unpopulated tanks in order to maintain system hygiene,
or it would need a large bypass valve that could be grad-
ually throttled as the system population grew.
Flow-Through Systems
While the vast majority of commercially available
zebrafish housing solutions employ recirculating aqua-
culture technology, there are applications when recy-
cling the culture water is not desirable. The application
of flow-through technology is the solution to this need.
In recirculating aquaculture systems, the culture water
used to house and grow the aquatic organisms is
captured, treated, and reused, thereby, greatly reducing
the total water volume consumed in the rearing process.
In flow-through or single-pass systems, new clean water
is delivered to the tanks, and the wastewater flowing out
of the tanks is rejected from the aquarium, typically to
the municipal sewerage system. For this reason,
flow-through aquarium systems are normally used
only for applications that require the complete isolation
of one tank from another.
294 26. Aquatic Housing

Toxicology and environmental manipulation experi- also needs to be applied as to whether there is need
ments are the archetypal applications that implement for additional special housing elements.
the single-pass system. With flow-through technology,
each tank or series of tanks can be exposed to different
chemical agents without contamination of others. Genotyping Racks
As the water exits the tank, it is routed to the sewer,
One of the most common activities that occurs in a
and thus, poses no cross-contamination risk to other
large-scale zebrafish production facility is the need to
tanks on the system. Flow-through technology is also
isolate individual fish while they are genotyped. Tradi-
appropriate to apply to high-risk quarantine scenarios
tionally this meant that fish would need to be main-
in which imported fish may be suspected of carrying a
tained in small static tanks for several days while their
disease that can have high pathogenicity and be difficult
genotype results were obtained. During this time the
to contain. Such a disease is likely to have a significant
fish cannot be fed, as uneaten food, and wastes released
detrimental impact on other fish residing under the
from the fish after eating, rapidly degrades the water
heightened biosecurity control of a quarantine aquar-
quality, even to levels that could result in death.
ium, and use of a flow-through system can greatly
Recognizing this technology limitation, some com-
reduce the dangers of cross-contamination that would
mercial manufacturers now offer specialized genotyp-
be present in a RAS.
ing racks in the approximate footprint of a typical
Aside from the increased operational cost associated
housing rack and designed to hold up to 288 fish in
with replacing 100% of the system’s water volume every
isolation while maintaining water circulation
pass, there is also a significant challenge associated with
(Fig. 26.13). This means the fish can continue to be fed
controlling the culture conditions of the incoming new
while in isolation, addressing the first of the five free-
water. In a RAS aquarium, it is logical to choose demin-
doms. This is a significant enhancement in the animal
eralized water as the source water for the aquarium,
welfare associated with the challenge of genotyping.
giving the operator complete control over the fish ready
water that circulates through the aquarium. The oper-
ator simply changes the chemical mix added into the Automated Heat Shock Systems
makeup water to alter the culture condition. In a flow-
through system, manipulating the water chemistry to Heat shock is a powerful technique for inducible gene
this extreme would be prohibitively expensive. As a expression systems, and the zebrafish is well suited to
result, most flow-through aquaria need to accept that exploit the system, with the existence of over 1300 trans-
the source water will largely determine the culture genic fish lines carrying a construct under the control of
condition. For most operators, this means being limited the heat shock promoter 70 (hsp70; Zfin transgenic data-
to using the municipal water supply, and applying basic base; http://zfin.org/). Heat shock treatment of devel-
treatment and disinfection to the incoming water on its oping zebrafish embryos is relatively simple to
way to the header pipes. As zebrafish are tolerant of a perform because the fish are largely stationary, but
wide range of culture conditions, having the basic water more importantly, they do not require water circulation.
quality dictated by the municipal authorities is not a Heat shock treatment of adult fish is more involved as
major issue provided the operator is aware of the poten- there is a need to maintain circulation as their culture
tial for fluctuation in the water chemistry as the seasons medium is raised to the necessary shock temperature.
change. In some cases, a facility may be able to utilize a
natural water source, such as a local stream or bore;
however, the basic water chemistry of the source water
will still be the defining culture condition of the system.
As a minimum, the incoming water should pass through
an activated carbon filter, to strip heavy metals and
chloramines from the water, and then pass through a
UV disinfection chamber to knock down potential
pathogens residing in the water source.

Advanced Applications

Beyond the major decisions surrounding the core FIGURE 26.13 Genotyping Rack with under-rack sump. Photo
housing solutions for a zebrafish facility, consideration courtesy of Tecniplast Aquatic Solutions.

III. Husbandry
 Advanced Applications
While this challenge can be addressed by simply trans-
ferring the fish from housing tanks to treatment tanks,
this approach requires significant handling, which is
labor-intensive and increases stress levels in the fish.
In larger recirculating systems, when only a few tanks
of adult zebrafish require regular heat shocking, it is
possible to place in-tank heaters with activation timers
into each of the required tanks to avoid the need for
excessive handling (Duszynski, Topczewski, & LeClair,
2011). This technique, however, also requires regular
calibration of equipment, and strict monitoring of treat-
ment conditions to ensure that tanks are exposed to
statistically equivalent conditions. While this works
well in treating a handful of tanks for short-term exper-
iments, if a more pronounced commitment to heat shock
techniques is desired, the major limitation of the RAS
system is that it is not amenable to scalability.
While, currently, there are no off-the-shelf, scalable
solutions to heat shocking adult zebrafish, it is possible
to modify existing aquarium housing designs to enable
heat shock treatment in conditions of continuous water
flow, without impacting the standard culture condition
295
of the main circulation loop. By adding an additional
manifold to an aquarium holding rack, and by modi-
fying the wastewater return path to circulate water
from a treatment water bath, it is possible to deliver
heat shock conditions to an entire row of tanks with
precision and minimal variability between treatment
tanks (Saera-Vila, Kish, & Kahana, 2015). This system
can be easily scaled to suit demand and offers greater
assurance of maintaining optimal culture condition
throughout the RAS aquarium. In addition, this design
modification, illustrated in Fig. 26.14, can be further
enhanced through the use of solenoid valves to control
manifold change over, offering a fully automated heat
shock solution that avoids unnecessary handling stress,
and ensures best water quality and fish welfare at all
times.
Introduction to Defined Flora
Zebrafish are also breaking new ground in the realm
of gnotobiology. The term “gnotobiology” (gnos, known;
bios, life) is used to describe the study of animals raised

FIGURE 26.14 Heat Shock modification of a multilink zebrafish housing rack. The recirculating heat shock (35
C) system (shown in red)
resides under the housing rack. A small pump circulates hot water to an additional water distribution manifold, and a secondary wastewater drain
returns the effluent water to the heat shock water bath, closing the loop. The manifolds and drain of the regular (28
C) system are shown in blue.
For exposing the fish to the treatment, the manifold isolation valves are switched (closing the regular manifold valve, and opening the heat shock
manifold valve) and the heat shock circulation pump is turned on. The main system drain of the gutter of the heat shock tanks’ shelf is closed,
diverting the heated effluent water into the extra heat shock drain, returning the water to the heat shock water bath, avoiding heat contamination
to the rest of the system. On conclusion of the treatment, the heat shock circulation pump is turned off, the main system drain is opened, and the
manifold isolation valves are returned to their normal state, allowing normal system water to mix with the heated water in the tanks, gradually
returning holding conditions to normal.

III. Husbandry
296 26. Aquatic Housing

in the absence of microorganisms or in the presence of
known microbial strains or communities (Reyniers
et al., 1949). Just as a geneticist performs genetic manip-
ulations to study loss and gain of function, gnotobiolo-
gists test the necessity of microbiota by rearing
animals in the absence of microorganisms and then
adding one or more defined microbial strains to test
sufficiency (Melancon et al., 2017). The first concept of
studying germ-free and defined flora animals began
back in 1885 when Louis Pasteur hypothesized the
impossibility of animal life surviving in the absence of
microorganisms. Since then, mammalian models have
dominated gnotobiology, with various fish models
playing only a minor role. The plethora of attributes
that make the zebrafish model system popular with
geneticists, including rapid external development, opti-
cal transparency, large brood size and ease of housing,
are also the key attributes propelling zebrafish as a
gnotobiotic model. In fact, the use of zebrafish is
enabling vertebrate gnotobiotic experimentation on a
scale not possible with the traditional mammalian
models (Melancon et al., 2017).
The vast majority of gnotobiotic studies employing
zebrafish to date have focused on the use of larvae, in
part due to the challenges of long-term germ-free zebra-
fish husbandry methods. For successfully administering
long-term gnotobiotic zebrafish husbandry requires
consideration of animal housing, water quality, and
exchange, waste removal, nutritional value of food,
live food culturing, work flows and space requirements,
record keeping, supplies, expenses, and labor. Proced-
ures for long-term gnotobiotic husbandry are evolving.
Melancon et al. (2017) have reported success in raising
germ-free zebrafish for up to 1 month; however, the
process was extremely labor-intensive. For more infor-
mation on techniques, please also refer to the chapter
in this book on gnotobiotics.
Presently, two general housing systems can be
employed to raise gnotobiotic zebrafish. Germ-free
zebrafish can be maintained outside of a gnotobiotic
isolator by rearing the fish in sterile tissue culture flasks
or multiwell plates within a biosafety cabinet, using
standard sterile techniques (Melancon et al., 2017).
Each flask or well can represent a different microbial
condition, allowing multiple microbiotic conditions to
be studied in a single experiment (Pham, Kanther,
Semova, & Rawls, 2008). This brings gnotobiotic experi-
mentation within the grasp of most investigators work-
ing with the zebrafish model. However, scalability of
growing gnotobiotic zebrafish in flasks has significant
limitations. The daily requirement to service these
experimental housing tanks increases husbandry labor,
which, in turn, increases the opportunity for contamina-
tion; this can become financially limiting as throughput
and longevity increase. Culture flasks and plates are,
therefore, ideal for experiments that require analysis of
multiple different microbial conditions with relatively
few subject animals per condition and/or can be
completed within a short time course with minimal
manipulations to reduce the risk of contamination
(Pham et al., 2008).
The second and more rigorous method of rearing
germ-free zebrafish is to rear them within a purpose-
built gnotobiotic isolator as described by Pham et al.

FIGURE 26.15 Schematic diagram of isolator equipped for gnotobiotic zebrafish husbandry. The gnotobiotic isolator allows zebrafish to be
reared for extended periods in a defined flora environment. The gnotobiotic isolator includes attached gloves, allowing sterile interaction with the
inhabitants (1). A heating pad or heated water pump located directly below the specimen holding tanks allows control of isolator operating
temperature (2). Foil-covered beakers containing zebrafish in gnotobiotic zebrafish medium (GZM) act as static specimen holding tanks (3).
Carboy of sterile GZM allows routine water changes in holding tanks (4), and a second carboy collects the wastewater. Bottles of sterile zebrafish
food (5), and other supplies (6) are also stored within the isolator. For maintaining a sterile environment, the isolator is kept under positive
pressure, with HEPA-filtered air supply (7). Air exhaust is likewise typically HEPA-filtered (8). The gnotobiotic isolator also has a sealed port to
enable the transfer of materials to and from the sterile enclosure (9). Figure adapted from Pham, L. N., Kanther, M., Semova, I., Rawls, J. F. (2008).
Methods for generating and colonizing gnotobiotic zebrafish. Nature Protocols, 3(12), 1862.

III. Husbandry
 References
(2008). In this work, fish were kept in a flexible film
isolator maintained under positive pressure and
supplied with high-efficiency particulate air (HEPA)
filtered air (Fig. 26.15). This device allows sterilized
food, water, and supplies to be introduced into the
isolator via a sealed port, and manipulations are per-
formed using attached gloves. Providing the gnotobiotic
isolator is of sufficient size, large quantities of sterile
food, water and supplies can be maintained within the
gnotobiotic environment, reducing the need to open
the isolator port to the external environment, and
thereby, reducing the risk of isolator contamination.
A single isolator can house a relatively large number
of zebrafish, but the limitation of this device is that all
animals within the isolator are subject to the same micro-
bial condition. In a typical experiment utilizing a gnoto-
biotic isolator, it is possible to house 360 germ-free
zebrafish embryos for a single experiment by distrib-
uting them into 12 400 mL, foil-covered glass beakers,
each containing 30 fish in 100 mL of gnotobiotic zebra-
fish medium (water) (Pham et al., 2008).
Acquisition of gnotobiotic isolators requires a signifi-
cant initial financial investment, and their routine main-
tenance is relatively laborious; thus, gnotobiotic
isolators are best suited to experiments that require
gnotobiotic animals be reared for extended periods of
time or for experiments that require large numbers of
animals to be exposed to the same microbial condition
(Pham et al., 2008).
As zebrafish gnotobiology expands, it is likely that
specialized aquarium systems will be developed to
allow the maintenance of gnotobiotic zebrafish
throughout their life cycle. Melancon et al. (2017) postu-
late that a combination of bioreactor-like stand-alone
aquarium systems, similar to vessels that have already
been developed for other fish (Forberg, Arukwe, & Vad-
stein, 2011), could be modified as follows for this pur-
pose. A system combining equipment that would
allow automated media changes, control of feeding
and colonization, and prevention of accumulation of
toxic byproducts together with interconnected mini-
bioreactor technology previously developed for
culturing microorganisms (Auchtung, Robinson, & Brit-
ton, 2015), might allow large-scale screening of gnotobi-
otic zebrafish. This could allow zebrafish to become the
only gnotobiotic vertebrate model for which large-scale,
long-term husbandry is possible.
Summary
It is an exciting time to be involved in the husbandry
of zebrafish. A steadfast and deliberate commitment for
the strategic development of housing solutions is
III. Husbandry
297
needed to support its continued development as a pre-
eminent model. The extent of research program support
will determine the scale and sophistication of the avail-
able solutions to be developed. While a simple box tank
on the bench may be sufficient to meet the needs of an
independent investigator seeking to address straightfor-
ward questions, a complex central life-support RAS
aquarium will allow hundreds to thousands of geneti-
cally unique lines to be maintained in a cost-effective
manner, supporting a number of investigators focused
on answering more complex problems. The experi-
mental application of the model will dictate the mate-
rials required to construct optimal primary and
secondary enclosures and the types of manufacturer
options needed for appropriate housing flexibility.
Specific tank design will depend on the goals of the
intended research programs utilizing the aquarium.
Housing and breeding for maintenance may be all that
is required for some research programs, but others
may require the production of vast quantities of
synchronized embryos on demand, an ability to monitor
physiological response to oxygen limitation, or the
effects of defined gut flora.
References
Adatto, I., Lawrence, C., Thompson, M., & Zon, L. I. (2011). A new sys-
tem for the rapid collection of large numbers of developmentally
staged zebrafish embryos. PLoS One, 6(6), e21715.
Auchtung, J. M., Robinson, C. D., & Britton, R. A. (2015). Cultivation of
stable, reproducible microbial communities from different fecal do-
nors using minibioreactor arrays (MBRAs). Microbiome, 3(1), 42.
Baensch, H. A., & Riehl, R. (1997). Aquarium atlas.
Best, J., Adatto, I., Cockington, J., James, A., & Lawrence, C. (2010).
A novel method for rearing first-feeding larval zebrafish: Polycul-
ture with type L saltwater rotifers (Brachionus plicatilis). Zebrafish,
7(3), 289e295. https://doi.org/10.1089/zeb.2010.0667.
Brand, M., Granato, M., & Nüsslein-Volhard, C. (2002). Keeping and
raising zebrafish. Zebrafish, 261, 7e37.
Brett, J. (1964). The respiratory metabolism and swimming perfor-
mance of young sockeye salmon. Journal of the Fisheries Board of
Canada, 21(5), 1183e1226.
Chen, S., Ling, J., & Blancheton, J.-P. (2006). Nitrification kinetics of bio-
film as affected by water quality factors. Aquacultural Engineering,
34(3), 179e197.
Clark, T. D., Sandblom, E., & Jutfelt, F. (2013). Aerobic scope measure-
ments of fishes in an era of climate change: Respirometry, relevance
and recommendations. Journal of Experimental Biology, 216(15),
2771e2782.
Collymore, C., Tolwani, R. J., & Rasmussen, S. (2015). The behavioral
effects of single housing and environmental enrichment on adult
zebrafish (Danio rerio). Journal of the American Association for Labora-
tory Animal Science, 54(3), 280e285.
Conradsen, C., Walker, J. A., Perna, C., & McGuigan, K. (2016). Repeat-
ability of locomotor performance and morphologyelocomotor per-
formance relationships. Journal of Experimental Biology, 219(18),
2888e2897.
Duszynski, R. J., Topczewski, J., & LeClair, E. E. (2011). Simple,
economical heat-shock devices for zebrafish housing racks.
Zebrafish, 8(4), 211e219.
298 26. Aquatic Housing
Ellis, T., North, B., Scott, A., Bromage, N., Porter, M., & Gadd, D. (2002).
The relationships between stocking density and welfare in farmed
rainbow trout. Journal of Fish Biology, 61(3), 493e531.
Forberg, T., Arukwe, A., & Vadstein, O. (2011). A protocol and cultiva-
tion system for gnotobiotic Atlantic cod larvae (Gadus morhua L.) as
a tool to study host microbe interactions. Aquaculture, 315(3),
222e227.
Fox, H. E., White, S. A., Kao, M. H., & Fernald, R. D. (1997). Stress and
dominance in a social fish. Journal of Neuroscience, 17(16),
6463e6469.
Harper, C., & Lawrence, C. (2011). The laboratory zebrafish. Crc Press.
Harris, J., & Bird, D. J. (2000). Modulation of the fish immune system by
hormones. Veterinary Immunology and Immunopathology, 77(3e4),
163e176.
Howdeshell, K. L., Peterman, P. H., Judy, B. M., Taylor, J. A.,
Orazio, C. E., Ruhlen, R. L., et al. (2003). Bisphenol A is released
from used polycarbonate animal cages into water at room
temperature. Environmental Health Perspectives, 111(9), 1180.
Kloas, W., Lutz, I., & Einspanier, R. (1999). Amphibians as a model to
study endocrine disruptors: II. Estrogenic activity of environmental
chemicals in vitro and in vivo. The Science of the Total Environment,
225(1), 59e68.
Larson, E. T., O’Malley, D. M., & Melloni, R. H. (2006). Aggression and
vasotocin are associated with dominantesubordinate relationships
in zebrafish. Behavioural Brain Research, 167(1), 94e102.
Lawrence, C. (2011). Advances in zebrafish husbandry and
management. Methods in Cell Biology, 104, 431e451.
Lawrence, C., & Mason, T. (2012). Zebrafish housing systems: A review
of basic operating principles and considerations for design and
functionality. ILAR Journal, 53(2), 179e191. https://doi.org/
10.1093/ilar.53.2.179.
Lawrence, C., Sanders, G. E., Varga, Z. M., Baumann, D. P.,
Freeman, A., Baur, B., et al. (2009). Regulatory compliance and
the zebrafish. Zebrafish, 6(4), 453e456.
Macdonald, N. P., Zhu, F., Hall, C., Reboud, J., Crosier, P., Patton, E.,
et al. (2016). Assessment of biocompatibility of 3D printed photo-
polymers using zebrafish embryo toxicity assays. Lab on a Chip,
16(2), 291e297.
Manuel, R., Gorissen, M., Stokkermans, M., Zethof, J., Ebbesson, L. O.,
Vis, H.v. d., et al. (2015). The effects of environmental enrichment
and age-related differences on inhibitory avoidance in zebrafish
(Danio rerio Hamilton). Zebrafish, 12(2), 152e165.
Matthews, J. L. (2004). Common diseases of laboratory zebrafish.
Methods in Cell Biology, 77, 617e643.
Maximino, C., De Brito, T. M., de Mattos Dias, C. A. G., Gouveia, A., &
Morato, S. (2010). Scototaxis as anxiety-like behavior in fish. Nature
Protocols, 5(2), 209e216.
McCarthy, I., Carter, C., & Houlihan, D. (1992). The effect of feeding hi-
erarchy on individual variability in daily feeding of rainbow trout,
Oncorhynchus mykiss (Walbaum). Journal of Fish Biology, 41(2),
257e263.
Melancon, E., Canny, S. G. D. L. T., Sichel, S., Kelly, M., Wiles, T.,
Rawls, J., et al. (2017). Best practices for germ-free derivation and
gnotobiotic zebrafish husbandry. Methods in Cell Biology, 138, 61e100.
Metcalfe, C. D., Metcalfe, T. L., Kiparissis, Y., Koenig, B. G., Khan, C.,
Hughes, R. J., et al. (2001). Estrogenic potency of chemicals detected
in sewage treatment plant effluents as determined by in vivo assays
with Japanese medaka (Oryzias latipes). Environmental Toxicology and
Chemistry, 20(2), 297e308.
III. Husbandry
Monstad-Rios, A. T., Watson, C. J., & Kwon, R. Y. (2017). ScreenCube: A
3D printed system for rapid and cost-effective chemical screening in adult
zebrafish (Zebrafish).
Mullins, M. C., Hammerschmidt, M., Haffter, P., & Nüsslein-
Volhard, C. (1994). Large-scale mutagenesis in the zebrafish: In
search of genes controlling development in a vertebrate. Current
Biology, 4(3), 189e202.
Parichy, D. M. (2015). Advancing biology through a deeper under-
standing of zebrafish ecology and evolution. Elife, 4, e05635.
Pauley, G. B., & Nakatani, R. E. (1967). Histopathology of “gas-bubble”
disease in salmon fingerlings. Journal of the Fisheries Board of Canada,
24(4), 867e871.
Pavlidis, M., Digka, N., Theodoridi, A., Campo, A., Barsakis, K.,
Skouradakis, G., et al. (2013). Husbandry of zebrafish, Danio rerio,
and the cortisol stress response. Zebrafish, 10(4), 524e531.
Pham, L. N., Kanther, M., Semova, I., & Rawls, J. F. (2008). Methods for
generating and colonizing gnotobiotic zebrafish. Nature Protocols,
3(12), 1862.
Procarione, L. S., Barry, T. P., & Malison, J. A. (1999). Effects of high
rearing densities and loading rates on the growth and stress re-
sponses of juvenile rainbow trout. North American Journal of
Aquaculture, 61(2), 91e96.
Reyniers, J., Trexler, P., Ervin, R., Wagner, M., Luckey, T., & Gordon, H.
(1949). The need for a unified terminology in germfree life studies.
Lobund Reports, 2, 151e162.
Saera-Vila, A., Kish, P. E., & Kahana, A. (2015). Automated scalable
heat shock modification for standard aquatic housing systems.
Zebrafish, 12(4), 312e314.
Schulte, C., & Nagel, R. (1994). Testing acute toxicity in the embryo of
zebrafish, brachydanio rerio, as an alternative to the acute fish test: Pre-
liminary results. Alternatives to laboratory animals. ATLA.
Sessa, A. K., White, R., Houvras, Y., Burke, C., Pugach, E., Baker, B.,
et al. (2008). The effect of a depth gradient on the mating behavior,
oviposition site preference, and embryo production in the zebra-
fish, Danio rerio. Zebrafish, 5(4), 335e339.
Sneddon, L. (2011). Pain perception in fish. Journal of Consciousness
Studies, 18(9e10), 209e229.
Spence, R., Magurran, A. E., & Smith, C. (2011). Spatial cognition in
zebrafish: The role of strain and rearing environment. Animal
Cognition, 14(4), 607e612.
Suomalainen, L. R., Tiirola, M., & Valtonen, E. (2005). Influence of rear-
ing conditions on Flavobacterium columnare infection of rainbow
trout, Oncorhynchus mykiss (Walbaum). Journal of Fish Diseases,
28(5), 271e277.
Traver, D., Winzeler, A., Stern, H. M., Mayhall, E. A., Langenau, D. M.,
Kutok, J. L., et al. (2004). Effects of lethal irradiation in zebrafish and
rescue by hematopoietic cell transplantation. Blood, 104(5),
1298e1305.
Varga, Z. M. (2011). Aquaculture and husbandry at the zebrafish interna-
tional resource center Methods in Cell Biology (Vol. 104, pp.
453e478). Elsevier.
Weber, D. N., & Ghorai, J. K. (2013). Experimental design affects social
behavior outcomes in adult zebrafish developmentally exposed to
lead. Zebrafish, 10(3), 294e302.
Wedemeyer, G. (1996). Physiology of fish in intensive culture systems.
Springer Science & Business Media.
Weitkamp, D. E., & Katz, M. (1980). A review of dissolved gas super-
saturation literature. Transactions of the American Fisheries Society,
109(6), 659e702.
Westerfield, M. (2000). The zebrafish book. A guide for the laboratory use of
zebrafish (Danio rerio) (4th ed.). Eugene: Univ. of Oregon Press.